This paper examines the effects of introducing compulsory attendance laws on the schooling of U.S. children for three overlapping time periods: 1880-1927, 1890-1927, and 1898-1927. The previous literature finds little effect of the laws, which is somewhat surprising given that the passage of these laws coincided with rising attendance. Using administrative panel data, this paper finds that laws passed after 1880 had significant effects on enrollment and attendance. Laws passed after 1890, for which both administrative and retrospective census data are available, had significant effects on enrollment, attendance, and educational outcomes. In both cases, the timing of increases in enrollment and attendance is consistent with a causal effect of the laws. For men in the 1898-1927 period who reported positive wage income in the 1940 census, compulsory attendance laws increased schooling and wage income. The OLS estimates of the return to a year of schooling are 8 percent and the IV estimates are 11 to 14 percent.